Build real software. Learn how it is really built.
Our internship is not a course with an internship attached. It is work, on real projects, inside a working method. That method is Ledger-Driven Development (LDD).
What makes this different
LDD treats a software spec as a ledger, not a cage. You are free to be creative, because every change is written down, classified and can be undone. Nothing you build is lost, and nothing is hidden.
A certificate alone does not get you hired. Being able to show what you built, and why you built it that way, does.
Every intern works inside the same record system you can see on this site: a plan, a journal, decisions, and a build spec that moves with the code.
Three ways to join
Paid Internship
You can contribute from day one, and Exon pays you. There is no course, so there is no course fee.
Assessment Phase
A one-month trial on a real project. No money in either direction. It ends with a viva — a spoken review of your work.
Internship with training
You do the internship while learning AI from the fundamentals. As you work for Exon you earn an allowance, and Exon pays the larger part of the course fee from it. See the internship policy for how the fee works.
Who should apply
This programme is built for people starting from the basics. If you have finished A/L and you are waiting to start a degree, looking for work, or you have only ever used AI like ChatGPT and want to understand how it really works, start here.
If you are already an undergraduate, a graduate, or self-employed, you are welcome to apply as well. We hold those applications for selection instead of starting them straight away, because the places we open first go to people beginning from the ground up. Your application stays on file — if a suitable place opens, we will call you.
What you need to apply
- Your first name and last name
- A WhatsApp number we can reach you on
- An email address
- Your district and city
- Which track you want, and a little about your background
By applying you agree to be contacted on WhatsApp. See the privacy notice.
Questions before you apply?
Message us on WhatsApp at 0777952396, or call 0788009955. Email [email protected].
Fee details depend on the track, so we go through them on a call rather than guess on a web page. Nothing about the fee is hidden — we simply want you to have the right numbers for your situation.
The method behind this programme
Ledger-Driven Development is published and citable. Dr. Ishantha Siribaddana, PhD, DBA, PGDip(IT), BSc(Chem Sp), SCJP — ORCID 0009-0002-7719-1282 · DOI 10.5281/zenodo.23218617.
Abstract
Spec-driven development promised to discipline AI-assisted software work: write the specification first, derive the code, keep the two in agreement. In practice the spec is abandoned within weeks — code drifts, the document becomes fiction, and every agent that later reads it builds on a lie.
This paper presents Ledger-Driven Development (LDD), a working model built and run at Exon Software Solutions. Its premise is a single principle: reversibility makes freedom safe. The specification is not a contract that forbids deviation; it is a ledger that records every deviation, classifies it, and makes it revertible. Creativity is unrestricted inside the sandbox — the guardrail is traceability, not restriction. Spec and code grow together, guarded by a journal, a deviation taxonomy, a commit convention, revert anchors, and one escalation gate that keeps directional changes with the project owner.
The model adds a governance layer that spec-driven development lacks: a ten-record system in which every working session writes an auditable trail — what happened, why, and how it was verified — so that a project built largely by AI agents remains reviewable by a human owner who did not watch every line.
The abstract above is quoted from the published paper, so it stays in English.
This is the abstract only. Read the full paper — the model, the deviation taxonomy, the record system and the figures — on Zenodo: https://zenodo.org/records/23218617.